Travel speed is a key measure of effectiveness for evaluating urban arterials. The methodology given in the Highway Capacity Manual requires specific signal parameters that are typically not available in planning phase and hence needed to be assumed. The use of assumed values of inputs that affect the results significantly raises methodological concerns. This paper demonstrated that a simple and practical method of estimating travel speed along urban arterial streets is possible and that such a method can be more accurate than the existing complex ones. Derived from the HCM delay formula, an analytical model to estimate speed on arterials, taking only inputs available to planners, is proposed. The calibrated model is evaluated and compared with the results obtained from the Highway Capacity Software package. Despite limited scope of inputs and simple structure of the model, the results match the simulated travel speeds closer than the HCS method. The presented research results clearly indicate a good possibility of developing a simple and practical method of predicting travel speed on urban streets for planning purposes.